Gene level expression versus variant-level expression

Question

[ramiro.barrantes]

We have tumor/normal sample pairs, did RNA sequencing and obtained gene level differential expression counts using star and salmon. We also used GATK ASEReadCounter to obtain allele-level counts for specific variants. I am looking for help in knowing how to interpret the difference in numbers.

For example,

Chr1:39302949 G->A we have:

• Gene Level differential expression counts: 8034 normal and 12000 in the tumor
• Allele Level count in the tumor: 4G and 6A
• Allele Level count in the normal: 48G and 42A

QUESTIONS:

1. On the one hand, I obtained this variant by sequencing the DNA of the same samples and doing variant calling using mutect, but it seems that the variant caller picked up an actual allele that exists in the population. How can that be?

2. How can I interpret the differences in the magnitude of the numbers between the gene and the allele level expression? If the gene is more highly expressed in the tumor than in the normal, could I also conclude that the variant also exists more frequently in the tumor?

Answer 1

[swbarnes2]

Only 10 counts in the tumor? I wouldn't conclude anything from that, except maybe there is a big drop in expression of transcripts with that exon in tumor versus normal.

[ramiro.barrantes]

Thank you for your response! Actually I think that this is nothing more than a heterozygous position, not really a mutation and that is what is being reflected here.